Transportation Research Record 1808 ■ 49

Paper No. 02-3171

State-of-Practice Review of
Deep Soil Mixing Techniques in China
Jie Han, Hong-Tao Zhou, and Fen Ye

Deep soil mixing techniques have been widely adopted in China since should be used for soft clay or silt having allowable bearing capaci-
the late 1970s and mainly used for ground improvement for structural ties less than 120 kPa. One of the most common applications of deep
support and earth-retaining structures for excavations. A comprehen- soil mixing is to support five- to nine-story buildings (2).
sive state-of-practice review on development, applications, soil-cement Since soil-cement mixtures have high compressive strengths and
properties, mechanistic studies, design methods, and quality control of low permeability, deep soil mixing techniques have their advan-
deep soil mixing techniques in China is presented. These research results tages in earth-retaining structures for excavations. Based on on-site
will serve as a reference for advancing deep soil mixing techniques in the conditions and the need for controlling lateral movements of walls,
United States and other countries. three soil-cement column wall systems are used in the practice:
unreinforced/unsupported soil-cement column walls, reinforced soil-
cement column walls, and composite soil-cement column walls. The
Deep soil mixing techniques were first introduced to China in 1977 unreinforced/unsupported soil-cement column walls are mainly used
and put into practical use in 1980 (1, 2). The early systems used for excavations with depths ranging from 3.0 to 6.0 m. The re-
cement slurry, which is referred to as the wet method. In 1983, an inforced soil-cement column walls and composite soil-cement col-
alternative system was developed to jet and mix lime powder with umn walls have been used for excavations with depths ranging from
soft soil. This system was quickly adopted later using cement pow- 6.0 to 12.0 m. Since tensile strengths of soil-cement mixtures are rel-
der, which is referred to as the dry method. Due to limited uses of atively low, they may not be strong enough to resist bending moments
lime for deep soil mixing in China, this article refers to soil-cement applied onto the columns by earth pressures. Steel beams (typically
column techniques as the deep soil mixing techniques. At the early H-beams) and bamboo have been inserted into columns to increase
stage, the deep soil mixing techniques were mainly used to improve tensile resistance. Bamboo was used in the early stage, whereas steel
soft ground for supporting structural foundations. Although the first beams have been in use since 1994. The steel beam-reinforced soil-
soil-cement column wall for excavations was constructed in 1981 cement walls typically require the soil-cement mixture to have a
and documented in a published paper in 1983 (3), significant devel- high cement/(wet) soil ratio of approximately 20% or more. To
opments of soil-cement column walls for excavations started at the reduce the cost of using steel beams, steel beams are commonly
end of the 1980s and the beginning of the 1990s after several suc- pulled out for reuse after the completion of projects. One apparent
cessful applications in soft soils, especially in Shanghai. Soil-cement benefit of using steel beams in soil-cement columns is the signifi-
columns can be arranged in isolated, grid, block, or continuous wall cant reduction of the sectional thickness of the wall. Gao et al. (7)
patterns. Because of concerns on excessive lateral movement and reported that a steel beam-reinforced soil-cement wall with 0.6-m
potential failure of deep excavations, soil-cement columns were re- sectional thickness was successfully used for supporting deep exca-
inforced with steel beams (typically H-beams) and bamboo or com- vations ranging from 9.7 to 10.7 m in soft clay. The soil-cement col-
bined with other supporting systems (such as soil nails, cast-in-place umn walls have also been used in combination with other supporting
bored concrete piles, and struts) to form composite wall systems. systems (such as soil nails, cast-in-place bored concrete piles, and
To provide guidance on design and construction of deep soil struts) to reduce lateral movements of walls for accommodating strict
mixing, several design and installation codes or specifications have deformation requirements.
been published, such as Technical Specification for Deep Soil Mixing
Method (YBJ225-91) (4), Soil Improvement Technical Code (JGJ 79-
91) (5), with a chapter on the deep soil mixing method, and Techni- ENGINEERING PROPERTIES OF
cal Specification for Retaining and Protection of Building Foundation SOIL-CEMENT MIXTURES
Excavations (JGJ 120-99) (6), with a chapter on soil-cement walls.
Numerous successful ground improvement projects for structural Unconfined Compressive Strength
support have been completed, which include buildings, storage or oil
tanks, highway embankments, and airfields. The main purposes of Unconfined compressive strength is one of the most important param-
using deep soil mixing are to increase bearing capacities and reduce eters used in the design and quality control of soil-cement columns.
total and differential settlements. The Soil Improvement Technical Numerous studies have been completed in China on unconfined
Code (JGJ 79-91) (5) recommended that deep soil mixing methods compressive strengths of soil-cement mixtures and their influence
factors in the last 20 years. As suggested in the Soil Improvement
Technical Code (5), unconfined compressive strengths can be deter-
J. Han, Department of Civil Engineering, Widener University, One University mined using cubic samples with a side length of 70.7 mm or 50 mm.
Place, Chester, PA 19013-5792. H.-T. Zhou, Shenzhen Geotechnical Investiga-
tion Institute, 15 Fuzhong East, Futian District, Shenzhen, Guandong 518026,
Moisture content of soil, cement type and content, curing condition
China. F. Ye, Department of Road and Traffic Engineering, Tongji University, and time, additives, and organic content are identified as key influence
Shanghai 200092, China. factors (8).
50 Paper No. 02- 3171 Transportation Research Record 1808

Moisture Content of Soil period criteria, which cause difficulties in design and construction,
researchers have sought for alternatives by establishing a correlation
Many tests have confirmed that a reduction in moisture content of between quick test results and 90-day test results. The explanation
soil increases the strength of the soil-cement mixture. Ten percent part of the Soil Improvement Technical Code (5) indicates that the
reduction in moisture content of soil yields a 10% to 50% increase unconfined compressive strengths of soil-cement samples cured for
in the unconfined compressive strength (9). 7 days and 30 days are about 30% to 50% and 60% to 75% of sam-
ples cured for 90 days, respectively. Based on experimental studies
on soft clay-cement mixtures in Shanghai, Liu (12) developed the
Type and Grade of Cement and Cement Content following correlation:

Although the most commonly used type of cement in deep soil mix- 0.42
quT 1 T 
ing is portland cement, other types of cements have been utilized also. =  1 (5)
quT 2  T2 
In Sun and Han’s study (10), slag cement showed different perfor-
mances than portland cement in increasing the unconfined compres-
where
sive strengths of soil-cement mixtures. For the same type of cement,
different grades of cements have also shown different performances quT1 and quT2 = laboratory unconfined compressive strengths of
in the strengths of mixtures. In China, cement is classified into six soil-cement samples cured for T1 and T2, respec-
grades: No. 275, No. 325, No. 425, No. 525, No. 625, and No. 725. tively, and
Cements with those grades have nominal compressive strengths of T1 and T2 = curing time.
mortar cubes of 1 part cement and 2.5 parts graded standard sand at
This correlation is based on the samples cured under water at a
27.0 MPa, 31.9 MPa, 41.7 MPa, 51.5 MPa, 61.3 MPa, and 71.7 MPa,
room temperature of 20°C for 7 to 90 days.
respectively, after a 28-day curing period (11). No. 425 and No. 525
Experimental studies also showed that the strengths of soil-cement
are the two common cement grades used in deep soil mixing so far.
mixtures depend on the conditions in which the samples are cured: in
Ye et al. (8) indicated that unconfined compressive strengths of soil-
air, under water, and in situ. Ye et al. (8) indicated that the strengths
cement mixtures with No. 525 grade cement are about 1.5 to 2.0 times
of soil-cement mixtures are affected by curing methods in the short
those with No. 425 grade cement.
term but not in the long term (90 days). However, the unconfined
Cement content in soil-cement mixtures has been defined in
compressive strengths of soil-cement samples from the field can be
three ways:
20% to 50% of samples from the laboratory (8).
Based on laboratory test results for soft clays from the southeast
weight of cement
α cw = × 100% (1) coast of China, Sun and Han (10) developed a regression equation
wet weight of soft soil with four key influence factors:
weight of cement
α cd = × 100% = (1 + wi )α cw (2) quT = 232.5 + (33.67α cd − 7.85wi + 7.59 I p ) ln T
dry weight of soft soil
(n = 33, R2 = 0.9) (6)
weight of cement
α cv = × 100% (3)
volume of soft soil where

where wi is initial moisture content of soft soil. quT = laboratory unconfined compressive strength (kPa) of a
It is suggested in the Soil Improvement Technical Code (5) that the soil-cement sample at the curing time, T,
cement content, αcw, should be used in the range of 7% to 15%. How- T = curing time (days), T ≥ 7 days,
ever, the later application of the steel beam-reinforced soil-cement αcd = cement content to dry weight of soft soil (%),
wall technology requires a higher cement content, αcw (approxi- wi = initial moisture content of soft soil (%), and
mately 20% or higher). On the basis of laboratory studies on soil- Ip = plastic index of soft soil.
cement samples with αcw = 5% to 20% at curing time from 0 to This formula was developed based on limited test data, specific
90 days, Liu (12) developed the following correlation: soil conditions, and No. 425 portland cement in China.

α 
1.77
qu1
=  cw1  ( 4)
qu 2  α cw 2  Additives

where qu1 and qu2 are the unconfined compressive strengths with A number of additives have been used in soil-cement mixtures for dif-
cement contents, αcw1 and αcw2, respectively. ferent purposes. For example, lignin CaCO3 has been used for water
reduction. However, test results showed it has an insignificant effect
on the strength of the mixture. The amount of lignin CaCO3 added in
Curing Time and Condition the mixture is typically 0.2% of the weight of cement. The addition
of gypsum or 3-aminoethnol or fly ash into mixtures was proved to
Many laboratory and field tests have demonstrated that the strengths increase the strengths of mixtures and reduce the quantity of cement
of soil-cement mixtures increase with the curing time due to hydra- needed. Typical amounts of gypsum and 3-aminoethnol used in mix-
tion, ion exchange between cement and clay, hardening, and solid- tures are 2% and 0.05% of the weight of cement, respectively. As pre-
ification. The Soil Improvement Technical Code (5) suggested that sented in the explanation part of the Soil Improvement Technical Code
the design strength should be based on the unconfined compressive (5), an addition of fly ash up to 100% of the weight of cement can
strength of a soil-cement sample cured for 90 days. Due to long test increase the strength of the mixture by approximately 10%.
 Han et al. Paper No. 02- 3171 51

Organic Content as the foundations formed by two distinct materials (columns and
the surrounding soil) having different moduli or stiffness, which
Organic content has an adverse effect on strengths of soil-cement mix- carry applied loads and deform together (15). These columns acting
tures. Zhou et al. (1) and Deng and Lin (13) demonstrated through as reinforcements can be soil-cement columns, stone columns, sand
experimental studies that the unconfined compressive strengths of piles, vibro-concrete columns, and so forth. Detailed discussion on
soil-cement mixtures were significantly reduced by increasing the characteristics of composite soil foundations can be found in the
organic content. In the study by Deng and Lin (13), unconfined com- literature (15, 16 ).
pressive strengths of soil-cement mixtures decreased by 57% when
the organic content increased from 1% to 3%.
Stress Distribution and Transfer

Cohesion and Friction Angle As a result of the modulus difference between soil-cement columns
and the surrounding soil, the stresses on the columns are different
Triaxial CU tests by Chao and Qian (14) indicated that the shear from those on the surrounding soil when the composite soil founda-
strengths of soil-cement mixtures increase with an increase of cement tion is subjected to the applied loads. This concept was confirmed by
content, in which the increase of cohesion is predominant. The cohe- many full-scale plate loading studies in the field that higher stresses
sion of soil-cement mixtures is about 10 times that of soft soil at αcw develop on soil-cement columns than those on the surrounding soil
= 7% and 20 times that of soil at αcw = 20%. Ye et al. (8) reported (15). Stress concentration ratio, defined as the ratio of the stress on
that the cohesion of soil-cement mixtures is about 20% to 30% of the column to that on the surrounding soil, has been widely used in
the unconfined compressive strengths. Friction angles of soft clay- China for investigating the stress transfer mechanisms of compos-
cement mixtures range from 20 to 30 degrees. The above cohesion ite soil foundations. Two sets of test data on stress concentration
and friction angles are expressed in a total stress concept. ratios are presented in Figure 2 as examples, one from a full-scale
plate loading study in the field (17 ) and another from a soil-cement
column-supported embankment (18). They both show that stress
Permeability concentration ratios increase with an increase of applied loads. The
finite-element study by Zhang et al. (19) confirmed this trend. After
As shown in Figure 1, the permeability of soil-cement mixtures
reaching a peak value in the plate load test, the stress concentration
depends on the initial permeability of soft soil and decreases with an
ratio drops due to yielding of the soil-cement column. Several studies
increase of cement content.
(20, 21) have shown that the stress concentration ratios for soil-
cement column-reinforced composite soil foundations mostly range
from 5.0 to 10.0 under design loads.
MECHANISTIC STUDIES

Ground Improvement for Structure Support

Excess Pore Water Pressures and
Composite Soil Foundations Rate of Consolidation

The terminology “composite soil foundations” or “composite ground” The influence of soil-cement columns on the generation and dissipa-
has been commonly used in the area of ground improvement in China tion of excess pore water pressures under applied loads has been
since late 1970s. Composite soil foundations are generally regarded studied by Zhang et al. (19, 22) using a finite-element method. In
their studies, the modular ratio, defined as the ratio of the elastic mod-
ulus of soil-cement columns to that of the surrounding soil, ranged
10 from 5.0 to 100.0. The analyses indicated that the influence depth of
Initial Soil Permeability excess pore water pressures in the surrounding soil was reduced by
Soil 1 (x10-5cm/s) an increase of the modular ratio. The relationships between accumu-
lated excess pore water pressures measured from plate loading tests
5.16
Permeability, k (x10-5 cm/s)

1 and applied loads for three cases were presented in Zhang et al. (19):
0.253
unreinforced, reinforced with stone columns, and reinforced with
0.0839
Soil 2 soil-cement columns. The results clearly showed that the generation
of excess pore water pressures was significantly reduced by the inclu-
0.1
sion of soil-cement columns. This conclusion was supported by Lin
Soil 3 and Wong (18) in their field embankment study.
In the consolidation study (22), two cases were selected for com-
0.01
parison: unreinforced and reinforced with soil-cement columns. Per-
meability of soil-cement columns and the surrounding soil was chosen
to be 1.0 × 10−8 cm/s and 1.0 × 10−7 cm/s, respectively. Elastic mod-
uli of soil-cement columns and the surrounding soil were 200 MPa
0.001 and 4 MPa, respectively. Area replacement ratio by soil-cement col-
0 5 10 15 20 25 umns was 30%. The computed degree of consolidation of soil under-
Cement Content, acw (%)
neath the footing at the depth equal to the radius of the footing was
from 10% to 40% for the unreinforced case, but 40% to 100% for the
FIGURE 1 Influence of cement content on permeability of soil- reinforced case. This study implies that the rate of consolidation can
cement mixtures (8). be accelerated by soil-cement columns due to the higher modular
52 Paper No. 02- 3171 Transportation Research Record 1808

14

12

Stress concentration ratio, n

10

8
Plate loading test (16)
6

4
Geosynthetic reinforced and
deep soil mixing columns
2 supported embankment (18)

0
0 100 200 300 400 500
Applied pressure, p(kPa)

FIGURE 2 Field data for stress concentration ratios of soil-cement columns.

ratio, although the permeability of soil-cement columns is even lower reported by Ye et al. (25) in Chinese and by Shao et al. (26) in Eng-
than that of the surrounding soil. The soil-cement columns with the lish. Typical strength properties of Shanghai clays are also included
higher modular ratio enhance load transfer from the soil to the in the figure. The figure indicates that a linear K0 = 0.7 earth pres-
columns. The contribution of the high modular ratio to the degree of sure distribution existed on both sides of the soil-cement wall before
consolidation due to the enhancement of load transfer was also proved the excavation. In the backcalculation of the K0 value, the total earth
in studies by Balaam and Booker (23) and Han and Ye (24) on stone pressure is defined as the effective earth pressure plus the water pres-
columns in addition to the drainage mechanism for stone columns. sure with a pressure coefficient of 1.0. During and after the excava-
tion, the lateral earth pressures above the base of the excavation (on
the retaining side) decreased and approached the active earth pres-
Earth-Retaining Structures for Excavations sures and those below the base of the excavation (also on the retain-
ing side) turned constant. The lateral earth pressures on the excavated
Earth Pressures (resisting) side decreased when the soil was removed and then
increased when the wall started to move. The simplified lateral earth
The variations of lateral earth pressures on retaining and excavated pressure distribution in Figure 3 has been commonly used as a basis
(resisting) sides of the soil-cement column wall were investigated in for development of theoretical solutions for stability and movement
a well-instrumented project as shown in Figure 3. These results were of soil-cement column walls.

Earth Pressure (kPa) Earth Pressure (kPa)

200 150 100 50 0 0 50 100 150 200 250
0 0
2 days before excavation Fill
2 10 days after excavation 2
Clay, ccu=15-18kPa
22 days after excavation φcu=17-200
4 4
38 days after excavation
Depth (m)

Silty clay
6 6 ccu=7-14kPa
φcu=13-180
8 8

10 10 Clay
ccu=9-12kPa
φcu=8-120
12 12

14 14
Simplified distribution

FIGURE 3 Measured lateral earth pressure distribution (25, 26).

Han et al. Paper No. 02- 3171 53

Residual Stresses f¯s = average skin friction between the column and the surround-
ing soil,
Most of the stresses below the base of excavation are released due to qt = end-bearing capacity, and
the removal of the soil. However, there exist residual stresses after the λ = bearing capacity reduction factor for end-bearing, ranging
removal of the soil. A residual stress coefficient was defined by Hou from 0.4 to 0.6.
and Liu (27) as the ratio of the residual stress to the unloaded stress.
The influence depth of residual stresses was defined as the residual The average skin friction is suggested in the Soil Improvement
stress coefficient equal to 0.95. Based on field data, the following Technical Code (5) based on the soil type as follows: Type I
correlation was established (27): (OC > 5%, e0 > 1.5), f¯s = 5 to 8 kPa; Type II (OC > 5%, 1.0 < e0 ≤
1.5), f¯s = 8 to 12 kPa; Type III (OC < 5%, e0 ≤ 1.0), f¯s = 12 to 15 kPa,
H where OC is the organic content and e0 is the in situ void ratio.
hr = ( 7) Zhou and Li (28) adopted the “pile compressibility” concept pro-
0.0612 H + 0.19
posed by Randolph and Wroth (29) to establish the relationship
where hr is influence depth of residual stresses (m), and H is depth between the relative stiffness K and the back-calculated bearing
of excavation (m). The residual stress coefficient is capacity reduction factor β from 13 field test data as follows (28):
β = 0.8 to 1.0 when K < 0.5; β = 0.5 to 0.8 when K = 0.5 to 0.8; β = 0.3
h2 to 0.5 when K = 0.8 to 1.0; β = 0 to 0.3 when K ≥ 1.0. The relative
χ = χ 0 + (0.95 − χ 0 ) 0 ≤ h ≤ hr (8)
hr2 stiffness K was defined as an inverse of the pile compressibility in
Randolph and Wroth (29), that is,
χ = 1.0 h > hr ( 9)
1 + νs  5 L(1 − νs )  E p d
where χ is residual stress coefficient, χ0 is residual stress at the base K = ln (13)
4  d  Es L
of excavation (χ0 = 0.30 for soft clay in Shanghai), and h is the depth
below the excavation base. where
K = relative stiffness of pile to soil,
DESIGN METHODS Ep = elastic modulus of the soil-cement column,
Es = elastic modulus of the surrounding soil,
Ground Improvement for Structural Support νs = Poisson’s ratio of the surrounding soil,
L = column length, and
Bearing Capacity d = column diameter.
Composite Bearing Capacity Following the Soil Improvement Bearing Capacity of an Equivalent Block The composite soil
Technical Code (5), the allowable design bearing capacity of a foundation formed by soil-cement columns and the surrounding soil
composite soil foundation can be estimated by can also be treated as an “equivalent block” or “equivalent pier.” The
edge of the equivalent block is defined as the outer edge of the cement-
Qc
qsc = m + β(1 − m)qs (10) soil columns. The applied pressure at the bottom of the equivalent
Ac block can be estimated by subtracting side resistance and surrounding
where soil resistance of the block from the total load (including the applied
load and the weight of the equivalent block) (1), that is,
qsc = allowable design bearing capacity of the composite soil
foundation; p0 A + W − fAs − qs ( A − A1 )
qs = allowable design bearing capacity of the surrounding soil; pz = ≤ qz (14)
A1
m = area replacement ratio;
β = bearing capacity reduction factor, 0.5 to 1.0 for friction where
columns and 0.1 to 0.4 for end-bearing columns;
Qc = allowable design load capacity of a single column; and pz = applied pressure at the bottom of the equivalent block,
Ac = cross section area of columns. p0 = applied pressure on the total loading area,
W = weight of the equivalent block,
The ultimate load capacity of a single column can be determined f = average skin friction between the equivalent block and the
from a field column load test or estimated as the minimal value from surrounding soil,
the following two equations: qs = allowable bearing capacity of the surrounding soil,
qz = allowable bearing capacity at the bottom of the equivalent
Qc = ηqu Ac (11) block,
As = side surface area of the equivalent block,
Qc = fsUp L + λ Ac qt (12)
A1 = section area of the equivalent block, and
where A= total loading area.

qu = unconfined compressive strength of the soil-cement sample

determined in the laboratory, Settlement
η = strength reduction factor of soil-cement in the field,
Up = perimeter of the column, The settlement of a footing on a composite soil foundation by soil-
L = column length, cement columns is considered as the sum of the compression of the
54 Paper No. 02- 3171 Transportation Research Record 1808

soil-cement column reinforced zone and the deformation of the soil ity, and global stability. For most applications of soil-cement column
underneath the reinforced zone. The formula for computing the defor- walls, the existing ground water levels are relatively high and there
mation of the reinforced zone (s1) is suggested in the explanation part exist differences of water levels between the retaining side and
of the code (5) as follows: the excavated side. The stability of walls due to seepage and /or
base heave should be checked also. The required minimal factors
( p0 + p0 z ) L of safety (FoS) are typically as follows (8): FoS (overturning) = 1.5,
s1 = (15)
2 E ps FoS (sliding) = 1.3, FoS (global) = 1.25, FoS (seepage) = 1.5, FoS
(base heave) = 1.2.
where
p0 = average applied pressure at the top of the reinforced zone;
p0z = additional stress at the bottom of the reinforced zone (This Embedment Depth of Wall
additional stress can be estimated by subtracting the over-
The technical specification (JGJ 120-99) (6) suggested that the
burden stress from the total vertical stress at the bottom of
embedment depth of a wall below the excavation base should be
the reinforced zone as expressed in Equation 14); determined using the Bishop circular slip surface method for global
L = length of soil-cement columns; and stability. When the excavation base is formed by permeable soils
Eps = composite constrained modulus of the reinforced zone, (gravel and sand), the embedment depth should be designed to satisfy
which can be estimated by the stability requirement for seepage. The minimal embedment depth
of the wall is 0.4 times the depth of excavation. This minimal embed-
E ps = mE p + (1 − m) Es (16) ment depth requirement was set based on the experience from a num-
ber of projects. In the explanation part of this specification, it is stated
or that the satisfaction of the requirement for the global stability can
ensure the embedment depth will meet the requirement for limiting
E ps = [1 + m(n − 1)]Es (17) heave of the excavation base. Based on the design experience in
Shanghai, the embedment depths of walls below the excavation base
where are in the order of 0.8 to 1.2 times the depth of excavation (32).
Es = constrained modulus of the soil determined from a consoli-
dation test;
Ep = elastic modulus of soil-cement determined from an uncon- Thickness of Wall
fined compression test [this modulus is defined as the ratio
The technical specification (JGJ 120-99) (6) suggested that the thick-
of the vertical stress reaching 50% of the unconfined com-
ness of a soil-cement column wall should be determined based on the
pressive strength to the vertical strain at the corresponding
stability against overturning. When the base of the wall is seated on
vertical stress; in absence of test data, this modulus can be
a gravel or sand layer, the water head difference between the retain-
estimated by (100 to 120) qu (8)]; and
ing and the excavated sides is considered as uplift water pressures
n = stress concentration ratio.
applied on the base of the wall. The minimal thickness of the wall can
In addition, the code (5) suggests that the deformation, s1, can be be estimated by
empirically estimated in the range of 10 to 30 mm based on the con-
ditions of applied pressures, column lengths, soil-cement strengths, 10(1.2 ha ∑ Fai − hp ∑ Fpj )
B≥ (18)
and so forth. 5γ cs ( H + hd ) − 2 γ w (2 H + 3hd − hwp − 2 hwa )
The method for computing the deformation of the soil underneath
the reinforced zone is suggested following another foundation code where
(30) for an unreinforced case. In this method, the soils within the influ-
ence depth are divided into many sublayers. The stress at the midpoint ΣF ai= total horizontal force acting on the back of the wall start-
of each sublayer is calculated using the Boussinesq solution. The set- ing from the wall base to the ground surface on the retain-
tlement of each sublayer is estimated based on the calculated stress, ing side;
the compression index, Cc /(1 + e0) of the soil, and the thickness of the ha = distance from the location of the total horizontal force,
ΣFai, to the base of the wall;
sublayer. The total settlement is the summation of the settlements
from all the sublayers. Σ Fpj = total horizontal resistance acting on the front of the wall
Han and Ye (31) proposed other settlement calculation methods from the wall base to the bottom of the pit at the exca-
considering reinforced and unreinforced zones as a double-layer vated side;
system having different moduli. hp = distance from the location of the total horizontal resistance,
ΣFpj, to the base of the wall;
γcs = average unit weight of soil-cement;
Earth-Retaining Structures for Excavations γw = unit weight of water;
hwa = depth of ground water from the ground surface on the
Stability Analysis retaining side;
hwp = depth of ground water from the excavation base on the
Soil-cement column walls are typically considered as gravity- excavated side;
retaining walls in stability analyses. Similar methods have been used hd = embedment depth of the wall;
as those for gravity-retaining walls to analyze the stability of soil- H = excavation depth; and
cement column walls, including overturning, sliding, bearing capac- B = width of the wall.
Han et al. Paper No. 02- 3171 55

hwa p = ky = (ξ z ) y (20)
γcs where
H
p = soil resistance;
ΣFai k = spring stiffness;
ξ = spring coefficient or subgrade reaction coefficient;
hwp ha y = the lateral displacement; and
hd ΣFpj z = the depth from the bottom of the pit.
hp
Field measurements verified an increase of the spring stiffness with
B
depth (25, 34). Based on the force diagram in Figure 5b, the force
equilibrium in the x-axis direction and the moment equilibrium yield
FIGURE 4 Definitions of dimensions and forces. the following equations:

24 M0′ − 8 H0′h 2 H0
y0 = + (21)
The definitions of the dimensions and forces are also presented in ξ h 3 + 36ξ I B ξh2
Figure 4. The above formula was developed based on a factor of
safety against overturning equal to 1.2. 36 M0′ − 12 H0′h
When the base of the wall is seated on a clay or silt layer, the θ0 = (22)
ξh 4 + 36ξ hI B
effect of the water head difference is ignored and Equation 17 can be
simplified as follows: where
y0 = lateral movement of the wall at the elevation of the exca-
2(1.2 ha ∑ Fai − hp ∑ Fpj ) vation base;
B≥ (19)
γ cs ( H + hd ) θ0 = rotation of the wall;
F h
This specification required a minimal wall thickness of 0.4 H, M′0 = M0 + H0 h + a1 − Mw ;
2
which was determined based on the experience from a number of M0 = moment on the wall at the elevation of the excavation base,
projects. Based on the design experience in Shanghai, the thickness computed based on earth pressure behind the wall;
of soil-cement column walls is in the order of 0.6 H to 0.8 H (33). B
Mw = W ;
2
Fa1 = active earth pressure below the pit base;
Lateral Movement H ′0 = H0 + Fa1 − SL;
H0 = horizontal force on the wall at the elevation of the excava-
Analytical Methods The approach proposed by Cai (32) has been
tion base, computed based on earth pressure behind the wall;
well accepted and used for estimating lateral movements of soil-
SL = sliding resistance along the wall base, which can be estimated
cement column retaining walls. In the derivation of his solution, the
by B  Su;
following assumptions were made to obtain a simplified model (Fig-
Su = undrained shear strength of soil;
ure 5): (a) the soil-cement column wall is rigid; (b) the earth pressures
B = width of the wall;
from the retained soil are assumed to have a linear K0 distribution to l  B4
IB = inertia of the wall section; I B = ; and
the depth of excavation and then turns constant down to the base of 12
l = unit length of the wall.
the wall; (c) the soil resistance on the excavated side of the wall is
modeled using springs with a linearly increasing stiffness with depth, Yang and Xiong (35) concluded from their study that the lateral
that is, movement of the soil-cement column wall also depended on the

B y0
y
W
M0
H H0

Fa1
h o
h/2
h
SL

e
(a) (b)

FIGURE 5 Simplified model for computing lateral movement of soil-cement

walls (32).
56 Paper No. 02- 3171 Transportation Research Record 1808

factors of safety against overturning, heave of pit base, and global reviewed. The article also presents the studies of lateral earth pres-
stability. sures and residual stresses after excavation. The design methods for
deep soil mixing techniques used for ground improvement and earth
Numerical Methods Numerical methods have been used for pre- retaining structures are summarized and discussed.
dicting and/or analyzing deformations and stability of soil-cement
column walls and base heave of excavation pits (26 ). It is a chal-
lenge to simulate the excavation sequence properly by considering REFERENCES
an unloading stress path. Proper determination of residual stresses
is also important to the accuracy of computation (27 ). 1. Zhou, G. J., T. Q. Lu, and E. X. Bai. Deep Mixing Method. In Soil
Improvement Handbook (in Chinese), China Architecture and Building
Press, Beijing, 1998.
2. Gong, X. N. Development of Deep Soil Mixing Methods in China (in
EVALUATION OF SOIL-CEMENT FOR Chinese). Proc., Symposium on Design and Construction of Deep Soil
QUALITY CONTROL Mixing Methods, China Railway Press, 1993, pp. 1–26.
3. Wu, T. A., X. G. Yang, and G. J. Zhou. Soil-Cement Retaining Walls
(in Chinese). Construction Techniques, 1983.
A number of testing methods have been adopted for evaluating the 4. Technical Specification for Deep Soil Mixing Method (YBJ225-91) (in
quality of soil-cement columns in the field, which include: the field Chinese). Building Research Institute, Ministry of Metallurgical Indus-
compression test, the dynamic penetration test, the static penetration try of China, 1991.
test, the sampling and laboratory compression test, the plate loading 5. Soil Improvement Technical Code (JGJ 79-91) (in Chinese). China
Building Research Institute, China Plan Press, 1992.
test, the refraction survey, the radar penetration test, and the wave 6. Technical Specification for Retaining and Protection of Building Foun-
velocity test. Advantages and disadvantages of each method were dis- dation Excavations (JGJ 120-99) (in Chinese). China Building Research
cussed in detail by Xu and Xu (36). Among all the methods, the first Institute, China Architecture and Building Press, 1999.
five methods are more commonly utilized. The field compression test 7. Gao, S. L., G. H. Yao, D. Y. Peng, and Y. K. Gui. Experimental Study on
Stress Behavior of Soil Mixing Walls (SMW method) (in Chinese). Proc.,
is used for evaluating the quality of soil-cement columns near the 8th Soil Mechanics and Geotechnical Engineering, Wanguo Academia
ground surface after the column heads are exposed. The dynamic pen- Press, 1999, pp. 651–654.
etration test (N63.5 with hammer mass of 63.5 kg or N10 with hammer 8. Ye, S. L., J. Han, and G. B. Ye. Soil Improvement and Underpinning
mass of 10 kg) and the static penetration test are used for examining Techniques (in Chinese), China Architecture and Building Press, Beijing,
China, 1994.
the uniformity of soil-cement columns with depth. These tests are
9. Wang, Y. L. Soil-Cement Mixing Techniques for Treating Soft Soil
mostly performed within a 7-day curing period. Empirical relation- Foundations (in Chinese). Internal report, the 1st Port Engineering
ships are used for estimating unconfined compressive strengths of Research Institute, Ministry of Transportation, 1979.
soil-cement mixtures (4, 8): 10. Sun, L. C., and J. Han. Influence Factors and Prediction of Unconfined
Compressive Strengths of Soil-Cement Mixtures (in Chinese). Journal
of Soil Improvement, No. 4, 1994, pp. 31–37.
qu = 12.5 N10 (23) 11. Zhang, N. L. Construction Materials for Roads (in Chinese). Tongji
University Press, Shanghai, China, 1992.
or 12. Liu, J. J. Experimental Study and Analysis of Engineering Properties of
Soil-Cement in Shanghai (in Chinese). M.S. thesis. Tongji University,
Shanghai, China, 1992.
1 13. Deng, J. T., and J. Lin. Influence of Organic on Soil-Cement Strengths
qu = qc (24) (in Chinese). Proc., Symposium on Design and Construction of Deep
10
Soil Mixing Methods, China Railway Press, 1993, pp. 50–53.
14. Chao, Z. K., and Y. L. Qian. Experimental Studies of Engineering Prop-
where erties of Soil-Cement Mixtures in Shanghai (in Chinese). Proc., 3rd Chi-
nese Conference on Soil Improvement, Qinhuangdao, China, Zhejiang
qu = unconfined compressive strength of soil-cement mixture University Press, 1992, pp. 148–153.
(kPa), 15. Han, J., and S. L. Ye. Analysis of Characteristics to Composite Grounds.
N10 = number of blow count, and Proc., Young Asian Geotechnical Engineers Conference, Bangkok,
qc = cone resistance (kPa). Thailand, 1991, pp. 197–206.
16. Ye, S. L., W. M. Cai, and J. Han. Soil Improvement by Deep Mixing
Core sampling is also utilized to exhume soil-cement samples Piles and Stone Columns. Proc., U.S.-China Workshop on Coopera-
tive Research in Geotechnical Engineering, Shanghai, China, 1992,
from soil-cement columns and then the samples are tested in the lab- pp. 57–73.
oratory for the unconfined compressive strengths. Plate loading tests 17. Gong, X. N. Composite Foundation (in Chinese). Zhejiang University
are commonly used for determining allowable bearing capacities of Press, 1992.
composite soil foundations and single soil-cement columns. 18. Lin, K. Q., and I. H. Wong. Use of Deep Cement Mixing to Reduce
Settlements at Bridge Approaches. Journal of Geotechnical and Geo-
environmental Engineering, ASCE, Vol. 125, No. 4, 1999, pp. 309–320.
19. Zhang, J., J. Han, and S. L. Ye. Analysis of Excess Pore Water Pressures
CONCLUSIONS in Soil-Cement Column Composite Foundations (in Chinese). Proc.,
Symposium on Design and Construction of Deep Soil Mixing Methods,
This article is a state-of-practice review of the use of deep soil mix- China Railway Press, 1993, pp. 168–172.
20. Qiu, Z. X., G. B. Ye, and D. S. Yang. Stress Concentration Ratios of
ing techniques for ground improvement and earth-retaining struc- Soil-Cement Column-Reinforced Composite Soil Foundations under
tures for excavations in China. Moisture content of soil, cement type Loading or Unloading (in Chinese). Proc., 3rd Chinese Conference on
and content, curing condition and time, additives, and organic con- Soil Improvement, Qinhuangdao, China, Zhejiang University Press,
tent have been identified as key influence factors for engineering 1992, pp. 174–178.
21. Zhang, X. N. Discussion on Stress Concentration Ratios of Deep Soil
properties of soil-cement mixtures. The research findings on stress Mixing Column-Reinforced Composite Soil Foundations (in Chinese).
distribution/transfer and excess pore water pressure/rate of consoli- Journal of Soil Improvement, China Soil Improvement Committee,
dation of composite soil foundations by soil-cement columns are Vol. 6, No. 2, 1995, pp. 9–14.
Han et al. Paper No. 02- 3171 57

22. Zhang, J., J. Han, and S. L. Ye. Analysis of Consolidation of Soil- 30. Foundation Design Code (GBJ 7-89) (in Chinese). China Building
Cement Column-Reinforced Composite Soil Foundations (in Chinese). Research Institute, China Plan Press, 1989.
Proc., 7th Chinese Conference on Soil Mechanics and Foundation Engi- 31. Han, J., and G. B. Ye. Design and Computation Methods for Double
neering, Chinese Society of Civil Engineering, China Architecture and Layer Systems in Ground Improvement (in Chinese). Final Report, Shang-
Building Press, Beijing, China, 1994, pp. 542–545. hai Municipal Construction Science Foundation for Young Investigators,
23. Balaam, N. P., and J. R. Booker. Analysis of Rigid Rafts Supported by Geotechnical Engineering Department, Tongji University, Shanghai,
Granular Piles. International Journal for Numerical and Analytical China, 1996.
Methods in Geomechanics, Vol. 5, No. 4, 1981, pp. 379–403. 32. Cai, W. M. Calculation of Horizontal Movement of Soil-Cement Column
24. Han, J., and S. L. Ye. Simplified Method for Consolidation Rate of Retaining Walls (in Chinese). Theory and Practice on Soft Soil Founda-
Stone Column Reinforced Foundations. Journal of Geotechnical and tion (D. Z. Gao ed.), China Architecture and Building Press and Tongji
Geoenvironmental Engineering, Vol. 127, No. 7, 2001, pp. 597–603. University Press, 1992, pp. 132–136.
25. Ye, B. L., M. S. Shi, and M. K. Xu. Research and Applications of Deep 33. Ye, S. L., and Y. S. Lou. Soil Improvement—General Report (in Chi-
Soil Mixing Retaining Structures (in Chinese). Proc., 3rd Chinese Con- nese). Proc., 7th Chinese Conference on Soil Mechanics and Foundation
ference on Soil Improvement, Qinhuangdao, China, Zhejiang University Engineering, Chinese Society of Civil Engineering, China Architecture
Press, 1992, pp. 27–32. and Building Press, Beijing, China, 1994, pp. 42–53.
26. Shao, Y., C. M. Zhang, and E. J. Macari. The Application of Deep Mix- 34. Wang, Y., and Z. L. Feng. Interaction between Bracing System and Soil
ing Pile Walls for Retaining Structures in Excavations. ASCE Geotech- During Deep Excavation. Proc., 9th International Conference on Com-
nical Special Publication No. 81: Soil Improvement for Big Digs. Proc., puter Methods and Advances in Geomechancis, Wuhan, China, A. A.
Sessions of Geo-Congress 98, Boston, Mass., 1998, pp. 84–95. Balkema/Rotterdam/Brookfield, 1997, pp. 2033–2036.
27. Hou, X. Y., and G. B. Liu. Estimation of Deformation due to Excava- 35. Yang, M., and J. H. Xiong. Analysis of Stability and Deformation of
tion (in Chinese). Proc., China Mainland-Taiwan Soil Mechanics and Soil-Cement Column Retaining Structures (in Chinese). Proc., 8th Chi-
Foundation Engineering Workshop, Xian, 1994, pp. 446–451. nese Conference on Soil Mechanics and Geotechnical Engineering,
28. Zhou, T. H., and M. S. Li. Design Methods and Coefficient Determina- Chinese Society of Civil Engineering, Wanguo Academia Press, 1999,
tion for Computing Bearing Capacities of High Strength Cement-Soil pp. 647–650.
Column Composite Foundations (in Chinese). Proc., 5th Chinese Soil 36. Xu, F. G., and F. Xu. Methods for Evaluating Quality of Soil-Cement
Improvement Conference, Wuyishan, China, China Architecture and Columns (in Chinese). Proc., 5th Chinese Soil Improvement Confer-
Building Press, 1997, pp. 233–237. ence, Wuyishan, China, China Architecture and Building Press, 1997,
29. Randolph, M. F., and C. P. Wroth. An Analysis of Deformation of Ver- pp. 320–323.
tically Loaded Piles. Journal of Geotechnical Engineering Division,
ASCE, Vol. 104, No. GT12, 1978, pp. 1465–1488. Publication of this paper sponsored by Committee on Transportation Earthworks.